IT people were definitely involved in the creation of these rules. I want to know what fantasy world you live in. In my experience, IT people are so rarely involved in policy at any given company that they might as well not exist.
That's because most perl coders never seem to learn top down design, instead slapping code down at random, wherever their cursor happens to be when they decide they need to add something.
don't mind lights that mean something useful. A nice moderately dim green LED that is only on when the system is on is fine. Changing it to red when the system is off is worthless, it tells me nothing that the absence of a lit LED would tell me.
It tells you the system is plugged in to a live outlet. My playstation isn't an electrical outlet tester. I'll rephrase though. It tells me nothing useful that the absence of a lit LED would tell me.
I don't mind lights that mean something useful. A nice moderately dim green LED that is only on when the system is on is fine. Changing it to red when the system is off is worthless, it tells me nothing that the absence of a lit LED would tell me.
Use of blue for power is not fine. Just because my speakers are on, they do not have to double as reading lights.
My router's light display is a mix of fine and not fine. The connection light that blinks for activity is useful, the lit logo (Linksys...er Cisco consumer grade unit) does not provide me with useful data except that "I finished booting". It could simply flash a few seconds and then go off and provide the same information, less waste.
My printer/scanner/fax unit has a reasonably dim green power LED, and a much bigger lit button of the current mode (copy/fax/scan buttons). Why not simply put that into the status display and light the button up only a few seconds? Lights need not be on all the time to give full information.
Even the cable modem lights, while useful, could be redone in a much more conservative manner with fewer LEDs, yet still communicate all the same information.
To those who like the christmas tree effect, go buy your case mod kids and USB light cords and don't ask others to hope they can properly adjust the monstrosities. I've used the tape solution, not to turn off the lights completely, but to dim them down enough so that when I'm sleeping in another room, I don't have enough light to read by. Before I did that, even with the door closed, I could still see the light coming out.
It isn't a simple binary choice of "no lights" or "useful lights". The problem is vendors deciding they like the new cheaper blue LEDs and putting them in without thinking about what the lights mean. Ignoring the old standards of what colors to use because now they have options that weren't available previously, or amping up the brightness not because they need to, but because they can. Some sanity could be restored to the issue.
You use Zones at home? What is the point of that? For the exact reasons that I would use zones at work, isolation. My web server and mail servers are now isolated from the key files, and a compromise of the web server doesn't automatically mean that the whole system is lost.
Solaris is only bearable to admin AFTER you replace all the archaic commandline tools w/ the more modern GNU equivalents . I've heard plenty of long time Solaris/HP-UX/AIX admins admit as much. I am a long time Solaris/HP-UX/Linux SA, and only rarely do I find myself wanting GNU tools that aren't part of the supported OS today. Yes, I like bash, but that's been a part of the OS for a few versions now, similarly with less. GNU grep? Never needed it. If it is that complex, a nawk script is usually better suited to the job. vim? Usually I'm fighting the idiotic defaults on the Linux boxes to turn off the extras that they try to turn on to be "helpful" to developers, but make my job harder (turning off colors in the ls command is the top complaint I get from other SAs who are scripting against Linux.)
There are of course add-ons that every OS needs to make it more usable, but I'd hardly call Solaris broken out of the box these days. In the old Solaris 2.6 days, yes. Solaris 8 went a long ways in correcting those issues, and it has been out long enough we are starting to worry about end of service life on it.
SA though is a lot more than just what utilities it has installed by default. Any SA worth their salt knows the POSIX tools enough to work on a base system. A lot of it is how the system is configured, items that can't be fixed by simply adding this GNUtool. What files to touch when adding a logical interface, the process of validating authentication mechanisms for users, volume management, performance management, etc. Linux has made tremendous strides forward in the realm of administration, but it is hardly the top OS in my experience. Every OS I've tried has had some serious breakage (the registry of AIX), or lots of little breakage (HP-UX). There are a few good things in Linux for SAs (/proc, logmon), but in their effort to improve some things, they actually made some of them worse (xinetd).
You can hardly fault GNU/Linux for improving on the old commercial *nix tools that have not changed significantly since the early 90's. Does it sometimes break old shell scripts? Sure. Does it frustrate longtime commercial *nix admins such as yourself? Definitly. But all in all it is an improvement IMO. If only Linux didn't feel the need to update a lot of the tools in ways to break things. xinetd? pam.d? passwd (the command)? I mentioned elsewhere how much I get bothered by the broken nature of the passwd -S flag on Linux which is a poor substitute for passwd -s on HP-UX, AIX, or Solaris. User review/auditing that actually works is just fundamental to SA.
Debian is definitely built for the sysadmin and Ubuntu is not meant for the server. Right Tool For The Job and all that. In the enterprise, I don't see true Debian, so I can't fairly comment on it. But I'm not talking about "does it require an SA to run, but is it SA friendly in a large heterogeneous environment?
What kind of scripting are you doing where you don't explicitly state which version of grep you will use? Isn't that covered in the second semester of basic scripting class? Never, call any external binary directly, always call a variable that explicitly declares which binary you are invoking.
As for the older tools, I've found that if I'm trying to do something complex enough that people are wanting to use GNU grep, usually they are doing it complex enough that I find it easier to go with nawk. Just because we use grep in our spoken language to each other doesn't make it the best possible tool. I can't count how many times I see grep a | grep -v b | grep c |... | grep... in scripts. Usually I can replace it with a single call to a short awk script.
I do a lot of cross platform scripting for security. I may bemoan the lack of a modern perl on all my HP-UX systems, but the fact that an OS doesn't come with the GNU versions of tools rarely bugs me.
Oh yeah, I shouldn't have to type "mount -F hsfs -o ro/dev/dsk/c0t6d0s2/cdrom" to mount a freaking cdrom and I shouldn't have to fiddle with vfstab or create my own little shell script alias for that. I should be able to just type "mount/cdrom" on a default install and have it work. It is those little types of polish that would make solaris a lot more usable. And here on my Sun systems, for years all I did was stick in the CD and it mounted. If for some reason it didn't mount, I'd just type `volcheck`. Don't run volfs? I can encode a mount alias in the vfstab just as easily on Solaris as I can on any other OS. A little easier actually because I can set the mount-at-boot flag to no on Solaris precisely for such things. Linux was always actually harder for me to get CD mounting to work if I shut down the volfs equivalence.
Solaris (or AIX or HP-UX:shudder:) is the OS for the data-center and things that are big. Linux is fine for most things that are mid-sized to smallish, and let's face it, the majority of things these days are mid-sized to smallish, despite what we may say when we're explaining things to management. If it's not maintaining over 20 concurrent Apache threads, or using a SGA of upwards of 8 GB Linux is fine. (and I'd bet the Apache thing would be fine there too).
Quit whining and learn UNIX. The greybeards are better to have on your side than to be fighting with. Okay, my beard doesn't have much gray in it yet, but I've found that Sun has made great strides on pushing down to the mid-size and smaller lately, especially with the T-2000 and zones. I forgot the RAM issues of the smaller hardware that you correctly pointed out. I'd also point out that part of what Sun sells is high performance hardware that looks like small hardware, but goes very fast, especially in RAM. HP-UX has some... quirks, but as much as some people hate it, I have a hard time complaining much. They've got a really intriguing concept in package management of the packages that are "configured" and then installed, which allows me to put package wrappers around tools that really, really want to be installed with standalone programs and not report through package management.
As for AIX, my normal line is "AIX may or may not be a good operating system. It just isn't Unix." This is usually followed by a few hours of cursing Linux because things that just plain work on AIX, HP-UX and Solaris need massive rethinking on Linux. Just when we think we have it figured out, a different operating system decides to be the freak and force us to do it differently. We've just gotten so used to AIX being the odd guy out that it no longer fazes us.
I'm someone who went from Linux to Solaris for home use. Admittedly, I have worked on Solaris for a few years, but the so called "Linux ease", is usually manifested as a series of one pain after another in documentation and scripting "All our major platforms do things this way, except Linux which does things this completely weird way." Even AIX (of AIX Isn't Unix fame), often did things closer to the Solaris way than the Linux way.
The big thing that converted me was zones. A lot of the other virtualizations take the approach of an entire separate computer and are based on the physical or logical domains that AIX and HP-UX have been pushing for years. In zones, I have a central global domain that can see, manipulate and help fix all the non-global zones. This allows me to administrate all the zones from the global zone, remembering I have one computer, yet the local zone isn't able to "break out" and see hardware. With zones, I put the app in the non-global zone and even if they get root, the idea is to prevent them from crashing the overall box or filling/rewriting the underlying hard drive.
At the time I did this, Xen wasn't mature enough to seriously use, it may have done some catch up by now, but I have stable zones that work well, and limit my external exposure.
NFS support. Linux's incredibly poor NFS server performance makes it unusable in my experience. Yes, that's with the kernel accelerator.
Easier sysadmin. I realize this is very subjective, but Linux always felt like they were trying to make the developer experience better, but in the world of SA, things just weren't up to the same level of the other Unixes. Gratuitous changes away from other other vendors did things get old real fast. Case in point, need to audit users and find out who has static passwords, who does not (we use a OTP system). All the other versions of Unix I work with have `logins -x` and `passwd -sa` except Linux. Their closest equivilence, passwd -S, uses a very different format and the results are actually unreliable or misleading.
Linux feels like it was made for developers and forgot system administrators. My scripts are a lot more complex any time I have to work on a Linux box. Most of my Linux experience is with Red Hat AS, I admit, maybe the more Debian systems remembered the SA, but somehow, I doubt it from what I've seen of Ubuntu. Don't use the GUI? You're back in a maze if twisty tunnels, no two alike.
Some of the newer features in Solaris that cause initial pain for the SA but really draw people in is the new integration of the rc?.d directories and inetd.conf into the svcs stuff. It feels like an unnecessary pain until you remember that if the daemon crashes for some reason, it will automatically try and restart, much like a high availability cluster. Solaris did appear to learn a lesson from AIX and kept the underlying files as XML which means I can parse them with my text editing tools reasonably.
The worst part of Solaris to me is the time to patch. Not the patch reporting tools, not the naming structure for patches, but the time for installation itself. Very slow.
Then again, HP-UX has an incomprehensible mechanism for patch obsoleting, that makes it very annoying to tell if one installed patch obsoletes another patch that isn't installed.
To me, the real question isn't "what does Solaris do so well", but "what does Linux do well enough to justify the pain of the special cases it requires to support?" The main answer seems to be Linux "runs on cheaper hardware", but that will really be hurt by the better virtualization tools like zones.
When they aren't too busy joining the Board of Education to make sure that no one gets the education they don't want their own kids to get. It seems more and more the people who run for school board are the ones who never sent their kids through the public schools.
Definitions vary too widely, and most porn meets the indecent, but not the obscene category. Even groups that operate above the level have indicated that they wish to retain their.com address. The issue is generally not such groups though as they already take active steps to avoid accidentally being hit by minors (registering proactively with censorware vendors, a warning page prior to entering the site proper, etc.)
Part of the problem is that if a site has multiple domains, how do you know that the given IP is crosslisted with a.XXX domain? Also, which country's standard should apply? Some European countries permit full frontal nudity on network television. The age of consent for nude modeling is not uniform. Would all nude modeling belong as porn? Even if it is not mandated from a legal viewpoint, there would still be the potential for tremendous pressure on websites that refuse to migrate over to a.XXX style domain.
RFC 3675 was referenced by someone else in this thread as a discussion point on this topic. I admit, I hadn't read it before, but a lot of it makes good sense to me. Translating from.kids to a.kids.us helps alleviate many of the problems of varying standards of child appropriate material from country to country.
The.XXX domain is an awful idea. If people want kid friendly content, then make a.kids or similar domain. There, membership can't be pushed on organizations that might be slightly on the shady side depending on which administration is in power, but sites that demonstrate a clear "kid friendly" content can opt in and it is easier to handle and with less potential for censorship of material available to adults.
Sometimes when I buy a CFL, I end up having to either return it or throw it away. Not because of a defect, but because it produces a flicker that can't quite be seen that induces migraines in some people. This sensitivity isn't all that rare, I've known others who had a similar problem.
I've been very reluctant to convert certain rooms of my house to CFL because of the fact that some CFLs give me problems after a few minutes, some take a few hours before the problem occurs. (Which is particularly important for my home office where I may work for ten or more hours at a stretch in the middle of the night while working a change.) Some models work well for me, some don't. My sensitivity is considered mild.
Incandescent lighting fixtures do make sense in some circumstances, even for home use.
As someone who has been a permanent telecommuter for the past several years, I disagree. But my entire company is diverse, geographically isolated. Working from an office didn't mean seeing anyone even in the same division I was in. Until very recently, no one in my state was on the same project. That wasn't a telecommuting result, that was the nature of the company. The project manager being in St Louis, the supervisor in New York, a few of the coworkers in Dallas (others in Michigan), and you quickly discover that it doesn't matter. When everyone is a telecommuter, no one gets penalized just for being a telecommuter.
Face time just isn't the issue these days, it is active communication. Most people I know who telecommute don't take the trouble to actively communicate with their peers and supervisor and thus fall behind for those reasons, their boss doesn't have that "warm fuzzy". The communication may be unrelated to work, but it has to happen. That is in fact, the main advice I give to remote workers, have the casual conversations with your peers even if it means calling them up and just chatting for ten minutes.
Also, not everyone is well suited for telecommuting. People who require more direct supervision to work well (we all know of people like that, they produce great work if the boss is there, but nothing when the boss is out) are poor candidates for remote working. Most times I see an example held up of telecommuting hurting someone's career, I find evidence that the individual either didn't know how to telecommute or wasn't well suited for it.
Derivation is irrelevent. Ease of use of the system for everyday use is the key for public adoption.
I don't care what system is used by scientists for their work. No one expects everyone to convert to base 16 counting because computer scientists find that a lot better to use for their work, so why would I care what scientists use for their job? Specialists will use specialized language and units for their job, it doesn't make those units appropriate for everyone.
Some of the metric units are fairly sane for everyday use, others are pretty impractical.
You complain about having multiple units for what one is trying to measure, it is precisely that lack that makes the metric system less convenient for my everyday use. I use teaspoons and tablespoons at the same time in cooking, then turn around and talk about cups. The dirty secret of the metric system is they do it too, they just try to pretend they are the same unit (kilometer vs meter vs centimeter). But divisions of ten aren't always good for normal use. It's good to have two units that differ by other factors, factors that came about because they were practical rather than because someone thought everything should be base ten. As people who are supposedly literate in computers, I would think that the slashdot community would be less addicted to counting on ten fingers.
Besides, "Just 5ml of sugar helps the medicine go down" doesn't have the same ring.
I would argue that metric has failed in the US because it just isn't any easier to use for the common person than the US system. What's particularly funny to me is the people who like to bring up units that just aren't used in common US usage as "examples" of the hideousness of the measurements used in the US. Yes, there are some metric units that by happenstance are easy enough to use to standard US units as to be convenient, but the meter/yard and quart/liter approximations are more accidents than the standard in my everyday life. I for one would rather think in the units that makes the numbers easier for me to use than try to use one unit that may or may not have easy to use numbers.
I also find that people tend to work better in fractions than in decimals, yet the metric system seems to have tried to exclude fractions. It suggests (though it doesn't have to be) imprecision, but it works great for everyday work.
My least favorite metric unit of all has to be Celsius. Even my science teachers (more than one) admitted that Celsius as a scale was not well thought out. Farenheit, despite the crazy arbitrariness, ended up being near perfect for talking about the weather. 100 is too hot to go out, 0 is too cold to go out.
Don't feel bad, the violas seemed to have the same, and they have the added insult of being almost no times that the viola gets highlighted for a phrase. At least in some pieces, the cellos get the melody.
Could be worse, could be Pachelbel's Canon, the same two measures literally repeated the whole piece.
When you move into a new residence, one of the first things most people do is order phone service. Phone companies often sell lists of this information, including name, address and telephone number. The way to handle this is when you get new phone service, tell the phone company you do not want to be on thist list. (Sorry, can't recall the formal name offhand.)
That kind of sorting isn't done anymore in many districts. Now you are sorted according to "Needs extra help to pass the test" and "can pass the test with ordinary work". In other words, remedial and everyone else. Note, there is no sorting based on being advanced, instead those students are expected to help teach the rest of the class and then sit silently as if it has some academic value to them.
Yes, this means putting a elementary school student working on multi-digit multiplication in the same math class as the student learning addition of single digits for the first time. High schools even are doing this a lot more for some reason that I cannot fathom. I sincerely hope this trend is limited to my area, but I can no longer presume it to be because of my conversations with other parents who are looking at changes in the past five years in the public school system in other districts.
The joke I used to hear was a guitar player can doodle but can't play what is written. A violin player can play what is written but can't doodle.
The main cases where I've seen guitar players improvising, a lot of times they just start playing semi random chord progressions, tweaking them to be in a complimentary (or same) key as others. I very rarely hear improvised melodies on guitar. Violin is rarely about chords, even when it is used to provide the harmony instead of the melody. Even now, most of the violin teachers I've been talking with don't really play chords, they fake them instead by playing two or more doublestops closely together. True chords seem to be rare there. On a guitar, the one semester I had of it in middle school, chords were taught almost as soon as the student could put their fingers on a dozen notes.
I'm not saying one is better than the other as both have their place. I don't want an individual player to improvise when I'm listening to an orchestra do Beethovan's first symphony. Other kinds of music it would be expected for the players to improvise.
Suzuki method which claims to be the most popular violin teaching method in the US, discourages strongly the improvisation skills. I also notice that violin students aren't taught jazz style playing, they are usually taught a collection of Baroque standards with one or two easier Mozart pieces tossed in once in a while. (It seems that Mozart had too big a violin when he was composing, it's the only reason I can think that he seemed to write so much for second position.)
So to me, it is still the teaching technique used that makes so much of a difference on some of these issues.
Re:Partial credit
on
The Expert Mind
·
· Score: 2, Interesting
After observing three different teachers closely and how they teach a young child in violin instruction, I can tell you even your basic premise is wrong that they will have the same "basics" level after a year.
Many Suzuki students after a year are completely unable to read music because they focus so much on rote play. Their technical ability in some areas may be very good, but they have not yet started on certain other areas. Usually these students are able to keep a beat, but their playing has always sounded mechanistic to me during the first two years. Also most Suzuki instruction I've observed has very little concern for correct hand hold and finger placement at least at first.
Another teacher I saw went rather slowly by Suzuki standards, not even starting Suzuki book one until the student had at least eight months instruction in the basics, but the student understood a surprising amount of theory with focus on music reading, notions of dynamics, musical phrasing, technical perfection etc. This student after a year won't know third position at all, but will have started exercises that will make it so when they do start doing shifts, their hand automatically does it correctly. But what they can play with that technical skill is likely to go very slow.
A third teacher focused heavily on making the piece musical, allowed some of the technical precision details to slip, etc. But the students of that teacher would be far more likely to be able to improvise or embellish, especially more than a Suzuki student. Their technical precision may be behind on finger placement, proper use of the "sweet spot" for placing the bow, etc. But they'll
A year just isn't near enough with a violin to get the basics. Three years seems more realistic because that's when the different teaching styles of what they emphasize seems to level out. But depending on the teaching style, some will be able to play by pure memory, some won't even be able to play anything by memory, and others will be able to play some by memory but will be able to improvise some. It really does matter in the teaching style.
When you set up phone service, the phone company puts your name/number on a list sold/rented to other companies for marketing of "newly established numbers." Historically you just had to specifically ask not to be put on the new phone service list to avoid the flood.
Found this out with an unlisted number that got massive numbers of calls from sales companies. Called the phone company to complain and was informed of the service. At least historically, the company tends to actually honor requests to not go on this list.
A word is defined as five characters (that includes the space) for purposes of wpm counting. But it also denotes text as opposed to say, code or random symbols. I type a lot faster when I'm typing words that I already know how to spell than I do when I'm typing in a sequence of letters that to me seems "random".
This rate is of course only useful when defined over at least a few minutes worth of typing. That is why good typing programs will list your sustained rate not just your burst rate (fastest typing speed).
So in a sense, yes, they are measuring cpm, except that it has a strong implication as to what kind of material is being typed. I never said it was a very useful number, but like many things, it is a bit of a standard that allows us to compare relative typing speeds of at least one kind.
Slashdot Mirror
That's because most perl coders never seem to learn top down design, instead slapping code down at random, wherever their cursor happens to be when they decide they need to add something.
It tells you the system is plugged in to a live outlet. My playstation isn't an electrical outlet tester. I'll rephrase though. It tells me nothing useful that the absence of a lit LED would tell me.
I don't mind lights that mean something useful. A nice moderately dim green LED that is only on when the system is on is fine. Changing it to red when the system is off is worthless, it tells me nothing that the absence of a lit LED would tell me.
Use of blue for power is not fine. Just because my speakers are on, they do not have to double as reading lights.
My router's light display is a mix of fine and not fine. The connection light that blinks for activity is useful, the lit logo (Linksys...er Cisco consumer grade unit) does not provide me with useful data except that "I finished booting". It could simply flash a few seconds and then go off and provide the same information, less waste.
My printer/scanner/fax unit has a reasonably dim green power LED, and a much bigger lit button of the current mode (copy/fax/scan buttons). Why not simply put that into the status display and light the button up only a few seconds? Lights need not be on all the time to give full information.
Even the cable modem lights, while useful, could be redone in a much more conservative manner with fewer LEDs, yet still communicate all the same information.
To those who like the christmas tree effect, go buy your case mod kids and USB light cords and don't ask others to hope they can properly adjust the monstrosities. I've used the tape solution, not to turn off the lights completely, but to dim them down enough so that when I'm sleeping in another room, I don't have enough light to read by. Before I did that, even with the door closed, I could still see the light coming out.
It isn't a simple binary choice of "no lights" or "useful lights". The problem is vendors deciding they like the new cheaper blue LEDs and putting them in without thinking about what the lights mean. Ignoring the old standards of what colors to use because now they have options that weren't available previously, or amping up the brightness not because they need to, but because they can. Some sanity could be restored to the issue.
There are of course add-ons that every OS needs to make it more usable, but I'd hardly call Solaris broken out of the box these days. In the old Solaris 2.6 days, yes. Solaris 8 went a long ways in correcting those issues, and it has been out long enough we are starting to worry about end of service life on it.
SA though is a lot more than just what utilities it has installed by default. Any SA worth their salt knows the POSIX tools enough to work on a base system. A lot of it is how the system is configured, items that can't be fixed by simply adding this GNUtool. What files to touch when adding a logical interface, the process of validating authentication mechanisms for users, volume management, performance management, etc. Linux has made tremendous strides forward in the realm of administration, but it is hardly the top OS in my experience. Every OS I've tried has had some serious breakage (the registry of AIX), or lots of little breakage (HP-UX). There are a few good things in Linux for SAs (/proc, logmon), but in their effort to improve some things, they actually made some of them worse (xinetd). You can hardly fault GNU/Linux for improving on the old commercial *nix tools that have not changed significantly since the early 90's. Does it sometimes break old shell scripts? Sure. Does it frustrate longtime commercial *nix admins such as yourself? Definitly. But all in all it is an improvement IMO. If only Linux didn't feel the need to update a lot of the tools in ways to break things. xinetd? pam.d? passwd (the command)? I mentioned elsewhere how much I get bothered by the broken nature of the passwd -S flag on Linux which is a poor substitute for passwd -s on HP-UX, AIX, or Solaris. User review/auditing that actually works is just fundamental to SA. Debian is definitely built for the sysadmin and Ubuntu is not meant for the server. Right Tool For The Job and all that. In the enterprise, I don't see true Debian, so I can't fairly comment on it. But I'm not talking about "does it require an SA to run, but is it SA friendly in a large heterogeneous environment?
What kind of scripting are you doing where you don't explicitly state which version of grep you will use? Isn't that covered in the second semester of basic scripting class? Never, call any external binary directly, always call a variable that explicitly declares which binary you are invoking.
... | grep ... in scripts. Usually I can replace it with a single call to a short awk script.
As for the older tools, I've found that if I'm trying to do something complex enough that people are wanting to use GNU grep, usually they are doing it complex enough that I find it easier to go with nawk. Just because we use grep in our spoken language to each other doesn't make it the best possible tool. I can't count how many times I see grep a | grep -v b | grep c |
I do a lot of cross platform scripting for security. I may bemoan the lack of a modern perl on all my HP-UX systems, but the fact that an OS doesn't come with the GNU versions of tools rarely bugs me.
Quit whining and learn UNIX. The greybeards are better to have on your side than to be fighting with. Okay, my beard doesn't have much gray in it yet, but I've found that Sun has made great strides on pushing down to the mid-size and smaller lately, especially with the T-2000 and zones. I forgot the RAM issues of the smaller hardware that you correctly pointed out. I'd also point out that part of what Sun sells is high performance hardware that looks like small hardware, but goes very fast, especially in RAM. HP-UX has some
As for AIX, my normal line is "AIX may or may not be a good operating system. It just isn't Unix." This is usually followed by a few hours of cursing Linux because things that just plain work on AIX, HP-UX and Solaris need massive rethinking on Linux. Just when we think we have it figured out, a different operating system decides to be the freak and force us to do it differently. We've just gotten so used to AIX being the odd guy out that it no longer fazes us.
I'm someone who went from Linux to Solaris for home use. Admittedly, I have worked on Solaris for a few years, but the so called "Linux ease", is usually manifested as a series of one pain after another in documentation and scripting "All our major platforms do things this way, except Linux which does things this completely weird way." Even AIX (of AIX Isn't Unix fame), often did things closer to the Solaris way than the Linux way.
The big thing that converted me was zones. A lot of the other virtualizations take the approach of an entire separate computer and are based on the physical or logical domains that AIX and HP-UX have been pushing for years. In zones, I have a central global domain that can see, manipulate and help fix all the non-global zones. This allows me to administrate all the zones from the global zone, remembering I have one computer, yet the local zone isn't able to "break out" and see hardware. With zones, I put the app in the non-global zone and even if they get root, the idea is to prevent them from crashing the overall box or filling/rewriting the underlying hard drive.
At the time I did this, Xen wasn't mature enough to seriously use, it may have done some catch up by now, but I have stable zones that work well, and limit my external exposure.
NFS support. Linux's incredibly poor NFS server performance makes it unusable in my experience. Yes, that's with the kernel accelerator.
Easier sysadmin. I realize this is very subjective, but Linux always felt like they were trying to make the developer experience better, but in the world of SA, things just weren't up to the same level of the other Unixes. Gratuitous changes away from other other vendors did things get old real fast. Case in point, need to audit users and find out who has static passwords, who does not (we use a OTP system). All the other versions of Unix I work with have `logins -x` and `passwd -sa` except Linux. Their closest equivilence, passwd -S, uses a very different format and the results are actually unreliable or misleading.
Linux feels like it was made for developers and forgot system administrators. My scripts are a lot more complex any time I have to work on a Linux box. Most of my Linux experience is with Red Hat AS, I admit, maybe the more Debian systems remembered the SA, but somehow, I doubt it from what I've seen of Ubuntu. Don't use the GUI? You're back in a maze if twisty tunnels, no two alike.
Some of the newer features in Solaris that cause initial pain for the SA but really draw people in is the new integration of the rc?.d directories and inetd.conf into the svcs stuff. It feels like an unnecessary pain until you remember that if the daemon crashes for some reason, it will automatically try and restart, much like a high availability cluster. Solaris did appear to learn a lesson from AIX and kept the underlying files as XML which means I can parse them with my text editing tools reasonably.
The worst part of Solaris to me is the time to patch. Not the patch reporting tools, not the naming structure for patches, but the time for installation itself. Very slow.
Then again, HP-UX has an incomprehensible mechanism for patch obsoleting, that makes it very annoying to tell if one installed patch obsoletes another patch that isn't installed.
To me, the real question isn't "what does Solaris do so well", but "what does Linux do well enough to justify the pain of the special cases it requires to support?" The main answer seems to be Linux "runs on cheaper hardware", but that will really be hurt by the better virtualization tools like zones.
When they aren't too busy joining the Board of Education to make sure that no one gets the education they don't want their own kids to get. It seems more and more the people who run for school board are the ones who never sent their kids through the public schools.
Definitions vary too widely, and most porn meets the indecent, but not the obscene category. Even groups that operate above the level have indicated that they wish to retain their .com address. The issue is generally not such groups though as they already take active steps to avoid accidentally being hit by minors (registering proactively with censorware vendors, a warning page prior to entering the site proper, etc.)
.XXX domain? Also, which country's standard should apply? Some European countries permit full frontal nudity on network television. The age of consent for nude modeling is not uniform. Would all nude modeling belong as porn? Even if it is not mandated from a legal viewpoint, there would still be the potential for tremendous pressure on websites that refuse to migrate over to a .XXX style domain.
.kids to a .kids.us helps alleviate many of the problems of varying standards of child appropriate material from country to country.
Part of the problem is that if a site has multiple domains, how do you know that the given IP is crosslisted with a
RFC 3675 was referenced by someone else in this thread as a discussion point on this topic. I admit, I hadn't read it before, but a lot of it makes good sense to me. Translating from
The .XXX domain is an awful idea. If people want kid friendly content, then make a .kids or similar domain. There, membership can't be pushed on organizations that might be slightly on the shady side depending on which administration is in power, but sites that demonstrate a clear "kid friendly" content can opt in and it is easier to handle and with less potential for censorship of material available to adults.
Sometimes when I buy a CFL, I end up having to either return it or throw it away. Not because of a defect, but because it produces a flicker that can't quite be seen that induces migraines in some people. This sensitivity isn't all that rare, I've known others who had a similar problem.
I've been very reluctant to convert certain rooms of my house to CFL because of the fact that some CFLs give me problems after a few minutes, some take a few hours before the problem occurs. (Which is particularly important for my home office where I may work for ten or more hours at a stretch in the middle of the night while working a change.) Some models work well for me, some don't. My sensitivity is considered mild.
Incandescent lighting fixtures do make sense in some circumstances, even for home use.
As someone who has been a permanent telecommuter for the past several years, I disagree. But my entire company is diverse, geographically isolated. Working from an office didn't mean seeing anyone even in the same division I was in. Until very recently, no one in my state was on the same project. That wasn't a telecommuting result, that was the nature of the company. The project manager being in St Louis, the supervisor in New York, a few of the coworkers in Dallas (others in Michigan), and you quickly discover that it doesn't matter. When everyone is a telecommuter, no one gets penalized just for being a telecommuter.
Face time just isn't the issue these days, it is active communication. Most people I know who telecommute don't take the trouble to actively communicate with their peers and supervisor and thus fall behind for those reasons, their boss doesn't have that "warm fuzzy". The communication may be unrelated to work, but it has to happen. That is in fact, the main advice I give to remote workers, have the casual conversations with your peers even if it means calling them up and just chatting for ten minutes.
Also, not everyone is well suited for telecommuting. People who require more direct supervision to work well (we all know of people like that, they produce great work if the boss is there, but nothing when the boss is out) are poor candidates for remote working. Most times I see an example held up of telecommuting hurting someone's career, I find evidence that the individual either didn't know how to telecommute or wasn't well suited for it.
Derivation is irrelevent. Ease of use of the system for everyday use is the key for public adoption.
I don't care what system is used by scientists for their work. No one expects everyone to convert to base 16 counting because computer scientists find that a lot better to use for their work, so why would I care what scientists use for their job? Specialists will use specialized language and units for their job, it doesn't make those units appropriate for everyone.
Some of the metric units are fairly sane for everyday use, others are pretty impractical.
You complain about having multiple units for what one is trying to measure, it is precisely that lack that makes the metric system less convenient for my everyday use. I use teaspoons and tablespoons at the same time in cooking, then turn around and talk about cups. The dirty secret of the metric system is they do it too, they just try to pretend they are the same unit (kilometer vs meter vs centimeter). But divisions of ten aren't always good for normal use. It's good to have two units that differ by other factors, factors that came about because they were practical rather than because someone thought everything should be base ten. As people who are supposedly literate in computers, I would think that the slashdot community would be less addicted to counting on ten fingers.
Besides, "Just 5ml of sugar helps the medicine go down" doesn't have the same ring.
I would argue that metric has failed in the US because it just isn't any easier to use for the common person than the US system. What's particularly funny to me is the people who like to bring up units that just aren't used in common US usage as "examples" of the hideousness of the measurements used in the US. Yes, there are some metric units that by happenstance are easy enough to use to standard US units as to be convenient, but the meter/yard and quart/liter approximations are more accidents than the standard in my everyday life. I for one would rather think in the units that makes the numbers easier for me to use than try to use one unit that may or may not have easy to use numbers.
I also find that people tend to work better in fractions than in decimals, yet the metric system seems to have tried to exclude fractions. It suggests (though it doesn't have to be) imprecision, but it works great for everyday work.
My least favorite metric unit of all has to be Celsius. Even my science teachers (more than one) admitted that Celsius as a scale was not well thought out. Farenheit, despite the crazy arbitrariness, ended up being near perfect for talking about the weather. 100 is too hot to go out, 0 is too cold to go out.
Don't feel bad, the violas seemed to have the same, and they have the added insult of being almost no times that the viola gets highlighted for a phrase. At least in some pieces, the cellos get the melody.
Could be worse, could be Pachelbel's Canon, the same two measures literally repeated the whole piece.
When you move into a new residence, one of the first things most people do is order phone service. Phone companies often sell lists of this information, including name, address and telephone number. The way to handle this is when you get new phone service, tell the phone company you do not want to be on thist list. (Sorry, can't recall the formal name offhand.)
ISO 8601 uses separators:
2001-09-11
It is in fact the type of separator used that helps identify the format used.
That kind of sorting isn't done anymore in many districts. Now you are sorted according to "Needs extra help to pass the test" and "can pass the test with ordinary work". In other words, remedial and everyone else. Note, there is no sorting based on being advanced, instead those students are expected to help teach the rest of the class and then sit silently as if it has some academic value to them.
Yes, this means putting a elementary school student working on multi-digit multiplication in the same math class as the student learning addition of single digits for the first time. High schools even are doing this a lot more for some reason that I cannot fathom. I sincerely hope this trend is limited to my area, but I can no longer presume it to be because of my conversations with other parents who are looking at changes in the past five years in the public school system in other districts.
Camera shops seem to be where I find the thumb drive sized card readers.
Some doctors would argue that people are tasting too much life these days, given our problems of obesity in the US.
The joke I used to hear was a guitar player can doodle but can't play what is written. A violin player can play what is written but can't doodle.
The main cases where I've seen guitar players improvising, a lot of times they just start playing semi random chord progressions, tweaking them to be in a complimentary (or same) key as others. I very rarely hear improvised melodies on guitar. Violin is rarely about chords, even when it is used to provide the harmony instead of the melody. Even now, most of the violin teachers I've been talking with don't really play chords, they fake them instead by playing two or more doublestops closely together. True chords seem to be rare there. On a guitar, the one semester I had of it in middle school, chords were taught almost as soon as the student could put their fingers on a dozen notes.
I'm not saying one is better than the other as both have their place. I don't want an individual player to improvise when I'm listening to an orchestra do Beethovan's first symphony. Other kinds of music it would be expected for the players to improvise.
Suzuki method which claims to be the most popular violin teaching method in the US, discourages strongly the improvisation skills. I also notice that violin students aren't taught jazz style playing, they are usually taught a collection of Baroque standards with one or two easier Mozart pieces tossed in once in a while. (It seems that Mozart had too big a violin when he was composing, it's the only reason I can think that he seemed to write so much for second position.)
So to me, it is still the teaching technique used that makes so much of a difference on some of these issues.
After observing three different teachers closely and how they teach a young child in violin instruction, I can tell you even your basic premise is wrong that they will have the same "basics" level after a year.
Many Suzuki students after a year are completely unable to read music because they focus so much on rote play. Their technical ability in some areas may be very good, but they have not yet started on certain other areas. Usually these students are able to keep a beat, but their playing has always sounded mechanistic to me during the first two years. Also most Suzuki instruction I've observed has very little concern for correct hand hold and finger placement at least at first.
Another teacher I saw went rather slowly by Suzuki standards, not even starting Suzuki book one until the student had at least eight months instruction in the basics, but the student understood a surprising amount of theory with focus on music reading, notions of dynamics, musical phrasing, technical perfection etc. This student after a year won't know third position at all, but will have started exercises that will make it so when they do start doing shifts, their hand automatically does it correctly. But what they can play with that technical skill is likely to go very slow.
A third teacher focused heavily on making the piece musical, allowed some of the technical precision details to slip, etc. But the students of that teacher would be far more likely to be able to improvise or embellish, especially more than a Suzuki student. Their technical precision may be behind on finger placement, proper use of the "sweet spot" for placing the bow, etc. But they'll
A year just isn't near enough with a violin to get the basics. Three years seems more realistic because that's when the different teaching styles of what they emphasize seems to level out. But depending on the teaching style, some will be able to play by pure memory, some won't even be able to play anything by memory, and others will be able to play some by memory but will be able to improvise some. It really does matter in the teaching style.
When you set up phone service, the phone company puts your name/number on a list sold/rented to other companies for marketing of "newly established numbers." Historically you just had to specifically ask not to be put on the new phone service list to avoid the flood.
Found this out with an unlisted number that got massive numbers of calls from sales companies. Called the phone company to complain and was informed of the service. At least historically, the company tends to actually honor requests to not go on this list.
A word is defined as five characters (that includes the space) for purposes of wpm counting. But it also denotes text as opposed to say, code or random symbols. I type a lot faster when I'm typing words that I already know how to spell than I do when I'm typing in a sequence of letters that to me seems "random".
This rate is of course only useful when defined over at least a few minutes worth of typing. That is why good typing programs will list your sustained rate not just your burst rate (fastest typing speed).
So in a sense, yes, they are measuring cpm, except that it has a strong implication as to what kind of material is being typed. I never said it was a very useful number, but like many things, it is a bit of a standard that allows us to compare relative typing speeds of at least one kind.